Genetic-Based Fuzzy Logic Controller for Satellites Stabilized by Reaction Wheels and Gravity Gradient

Abstract

In this paper, a systematic technique is proposed to design an optimal distributed fuzzy logic controller (FLC) for the attitude control of satellites stabilized by reaction wheels and gravity gradient. A single FLC with two inputs and one output is used for the pitch motion while a distributed controller with four FLCs is proposed for the roll and yaw motions to take into consideration the dynamic coupling between roll and yaw. The control action for yaw is determined by the yaw dynamics as well as the roll dynamics and similarly for the roll. The rules, the distribution of membership function of the FLC, and the linking parameters of the distributed controller are determined based on solving a constraint optimization problem using the genetic algorithms. To get accurate pointing, long-life time, and fast response for the satellite, the deviation of the satellite from its nominal position, the consumed power, and the time of deviation are included in the optimization objective function. With the use of multiple initial conditions in determining the objective function, the designed controller is shown to be stable with a satisfactory performance in a broad range of the operating conditions